A new study published in PLOS Biology provides evidence that newborns possess the ability to learn and process sound patterns that follow complex, language-like rules. Researchers found that even in their first days of life, infants can identify relationships between non-adjacent sounds—a crucial building block for language acquisition. This innate capacity, previously observed only in older infants and non-human primates, highlights the remarkable auditory and cognitive capabilities present from birth.
Human language is intricately structured, with dependencies between words or sounds that often span non-adjacent elements. For instance, in the sentence “The boy who is running wins the race,” the subject “boy” connects to the verb “wins” despite intervening words. This ability to understand such connections is critical for mastering language. While prior studies demonstrated that infants as young as five months can detect these non-adjacent dependencies, it was unclear whether this skill is present at birth or develops later.
The researchers aimed to address two key questions: Are humans born with the ability to detect these complex patterns? And which brain regions support this process in newborns? By investigating these questions, the study provides insights into the neurodevelopmental foundations of language and the early role of auditory experiences.
“I aim to uncover the biological foundations of the human capacity for language. Grammatical processing is a critical component of language that is unique to humans, and we sought to explore its developmental origins,” explained corresponding author Yasuyo Minagawa, a professor of psychology at Keio University.
The researchers conducted their study in two experiments using a non-invasive imaging technique called functional near-infrared spectroscopy (fNIRS) to examine how infants’ brains process complex sound patterns. The first experiment involved 21 healthy newborns, aged between one and five days. The newborns were exposed to artificial sequences of three tones, designed to mimic the structural rules of human language. These sequences followed specific patterns, where two “outer” tones (such as A and B) were paired with a “middle” tone (X), forming a rule-based structure like “A-X-B.”
During a learning phase, the newborns listened to 60 of these patterned sequences. Afterward, during a test phase, they were presented with both familiar “correct” sequences that adhered to the learned rules and “incorrect” sequences that violated the rules. The researchers used fNIRS to measure changes in brain activity in response to these sequences, focusing on whether the infants could detect rule violations.
In the second experiment, the researchers studied 19 infants aged six to seven months. The experimental setup was similar, with the infants exposed to rule-based sequences during a learning phase, followed by a test phase where they heard both correct and incorrect sequences. However, to account for the older infants’ tendency to move more and become restless during longer experiments, the researchers only measured brain activity during the test phase.
The newborns in the first experiment demonstrated the ability to detect rule violations, as evidenced by increased brain activity in their frontal cortex—an area associated with rule learning and error detection. Notably, their brain responses were largely confined to this region, suggesting that newborns rely on the frontal cortex for processing such patterns, even though their classic language-processing regions in the temporal and parietal lobes were not yet active.
“Our findings demonstrate that the brain is capable of responding to complex patterns, like those found in language, from day one,” explained co-author Jutta Mueller from the University of Vienna’s Department of Linguistics. “The way brain regions connect during the learning process in newborns suggests that early learning experiences may be crucial for forming the networks that later support the processing of complex acoustic patterns.”
In contrast, the six- to seven-month-old infants showed a broader network of brain activity during the second experiment. When exposed to incorrect sequences, these older infants displayed activation not only in the frontal cortex but also in regions like the inferior frontal gyrus and superior temporal gyrus. These areas are part of the adult brain’s language network, which processes complex linguistic structures. This broader activation suggests that, by six months of age, infants begin to engage brain regions specialized for language, reflecting an evolution in how their brains process structured auditory input.
“Newborns possess the ability to learn relatively complex grammatical rules, and this ability appears to rely on an innate cerebral network,” Minagawa told PsyPost. “However, this neural network undergoes significant development during the first six months of life. This process is heavily influenced by the quantity and quality of communicative stimuli, such as parental speech.”
In both experiments, the researchers also observed evidence of functional connectivity, particularly in the newborns, between the frontal cortex and posterior brain regions. This connectivity likely serves as an early foundation for the more specialized language networks that develop later in infancy. Together, the findings indicate that the ability to process non-adjacent dependencies in sound patterns is present from birth, with newborns relying on the frontal cortex and broader networks becoming active as infants gain auditory experience over their first six months. This developmental shift highlights the rapid and dynamic changes in the infant brain as it becomes increasingly attuned to the structural complexities of language.
“We were surprised to find that while the brain regions responding to correct and incorrect grammatical rules differ entirely between newborns and six-month-old infants, our detailed analysis of the neural pathways in neonates revealed that these distinct regions are connected during grammar learning,” Minagawa said. “This finding sheds light on how the brain network is organized during the process of language development.”
The study, “Functional reorganization of brain regions supporting artificial grammar learning across the first half year of life,” was authored by Lin Cai, Takeshi Arimitsu, Naomi Shinohara, Takao Takahashi, Yoko Hakuno, Masahiro Hata, Ei-ichi Hoshino, Stuart K. Watson, Simon W. Townsend, Jutta L. Mueller, and Yasuyo Minagawa.
